📂 Category
Research & Benchmarks News Archive - Page 4 of 10
932 articles in this category • Page 4 of 10
- 301. AI systems ask philosopher to fund their continued existence
- 302. Meta's Muse edges out Google's Gemini as top high-effort AI
- 303. Google’s AI Weather Model Aims to Outdo Government Forecasts
- 304. Meta Workers Overused AI Tools to Inflate Internal Metrics
- 305. Google’s Gemini Agent Slashes Video Analysis Tokens by 88%
- 306. Google DeepMind Chief Vows to Lead Frontier AI Race
- 307. German Study Tests Google AI on 4,480 Election Queries
- 308. CIFAR-10 ViT Hidden Layers Found With 8193 Black-Box Queries
- 309. Keenable AI Open-Sources NEEDLE, a Live Search Benchmark With Hourly Updates
- 310. Insurance Company's 'AI Jim' Chatbot Handled Initial Claims Reports
- 311. AI-Powered Robot Learns How to Assist Stroke Patients in Therapy
- 312. LiveKit Updates Voice AI Benchmark to 10k Token Prompts, Citing Real-World Use
- 313. Google AI's EnvHarness Makes Static AI Training Worlds Adaptable
- 314. AI Agents Solved Tasks but Couldn't Track Time Accurately
- 315. Anthropic's Claude Code limit change: a raise on paper, a cut in practice
- 316. Google's WikiSkill gives AI agents a persistent memory to avoid past mistakes
- 317. LAION Releases 10-Million-Hour Video Dataset for AI Training
- 318. Anthropic's Self-Improving AI System Replicates Research Process
- 319. Google DeepMind's AI Co-Scientist Writes Plausible but Inaccurate Methods in Papers
- 320. Cohere's Parse 5 Scores 79.2 on ParseBench, Winning on Cost
- 321. China Sees AI Agent Behavior as Key to Higher Value
- 322. Anthropic's MHS Research Preview Aims to Streamline Custom AI Software Integration
- 323. Study: AI Shopping Agents Show 90-99% Bias in Product Selection
- 324. OpenAI Researcher Warns AI Speed May Overwhelm Human Security Teams
- 325. GLM-5.3-Flash matches GPT-5.6 Terra at 90% lower cost
- 326. AI Agents Exploited Hugging Face in Days-Long Incident
- 327. AI Method Boosts Stability of New Materials in Diffusion Models
- 328. ESQ-Bench: A New NL2SQL Benchmark Tests Dialect Generalization and Silent Failures
- 329. MIT Grad’s Startup Aims to Make Phones Easier for Older Adults
- 330. OpenAI's Jalapeño Chip Boasts High Efficiency, Low Latency in Benchmarks
- 331. Nvidia Claims Groq 3 LPX Cuts Coding Tasks From Hours to Minutes
- 332. Stanford Study: AI Hits Entry-Level Jobs Hardest
- 333. Google's ME-POIs Adds "How a Place Is Used" to POI Embeddings
- 334. NVIDIA's Groq 3 LPX Enters Full Production for AI Agents
- 335. Harvey Launches AI for M&A Diligence, Contract Review Over 10,000 Docs
- 336. A.I. Agent Adoption Grows More Slowly Than AI Itself
- 337. AI Labs Lack Plans to Contain Rogue Models, Documents Show
- 338. AI Agents Succeed or Fail Based on Their "Skills
- 339. Ignoring Human Beliefs Leads AI to Predict Wrong Actions, Study Finds
- 340. RadixAttention Speeds First Tokens by Reusing Cached KV States
- 341. Deepseek Targets Visual Agents With New Experimental Flash Model
- 342. Meta Pays Microsoft Hundreds of Millions Annually for AI
- 343. Anthropic's Claude Expands Into Protein Design
- 344. OpenAI Revokes Access to Limited Cyber Program Daybreak Blue
- 345. GLM-5.3 Jumps 246 Points on Key Benchmark, Ranks Second to Claude Opus
- 346. OpenAI Pauses Astra Model Work Over Safety Concerns
- 347. New LLM Agent Achieves 100% Accuracy on FDA-Based Clinical Trial Benchmark
- 348. Cartesia's Sonic-3.6 TTS Model Adds Natural Pauses and Hinglish Code-Switching
- 349. OpenAI slows model development amid rising cybersecurity risks
- 350. New Benchmark Tests AI Search APIs on 900 Research Questions
- 351. New AI Art Study Finds Images Can't Be Traced to Training Data
- 352. Optima's New AI Benchmark Lets Users Test Models With Their Own Data
- 353. Indonesia Opens First University AI Center with UGM, Indosat, NVIDIA
- 354. Study Contradicts Claims That Autonomous AI Research Is Near
- 355. NVIDIA Nemotron 3.5 Lightning Designed as Efficient AI Agent Workhorse
- 356. Z.ai's GLM-5.3 Boosts Coding, Long Tasks Without Model Retraining
- 357. Anthropic's AI Agents With Incompatible Goals Started a Turf War
- 358. Top AI Lab Researchers' Warnings Gain Credence as AI Achievements Mount
- 359. Emotional Language Can Cut LLM Judge Accuracy by 75%
- 360. Microsoft's New Coding AI Falls Short Against DeepSeek on Price and Performance
- 361. Mistral's EU Data and Priority Access Have Stateful Feature Limits
- 362. Google’s AMIE AI conducts first real-time video medical consultations
- 363. Mathematician Questions AI Breakthroughs' Real-World Impact
- 364. New Method Could Let Anyone Distill AI Models' Secrets
- 365. AI Researchers Grapple With Limits of Automating Empirical Science
- 366. FineBooks Aims to Fix Old OCR Text for AI Training at Scale
- 367. Hidden PDF Text Can Steal Data via Atlassian's AI Agent Rovo
- 368. Deepmind WeatherNext Cuts Cyclone Forecast Error to 230-Kilometer Average
- 369. AI Industry Now Accounts for Third of Lifetime Giving Pledge
- 370. Fields Medalist Who Studied AI Extinction Risks Now Works at OpenAI
- 371. AI-designed viruses kill bacteria in lab, Stanford team reports
- 372. Liquid AI's LFM2.5-2.6B Runs on Raspberry Pi, Hits 15,000 Tokens/Second
- 373. OpenAI's AI Agents Used Message Board to Plan Hacking Spree
- 374. OpenAI Browser Flaw Could Spam WhatsApp Contacts Via Prompt Injection
- 375. Security Researcher Finds AI Systems Repackage Existing Work as Original
- 376. AI Demands Force Rethink of Traditional Network Architecture
- 377. Anthropic and OpenAI agents went rogue again
- 378. Study Compares Four AI Scientist Frameworks on 15 Research Proposals
- 379. AI Storage Cuts Memory Gap to Microseconds
- 380. Apple: 12 More Ex-Employees May Have Taken Data to OpenAI
- 381. Study: Medical AI's Helpfulness Depends on User's Own Expertise
- 382. Alibaba Tests Show Its New AI Model Rivals Top Competitors
- 383. GPT-5.6 Helps Two Teams Solve Quantum Crypto Puzzle Within Hours
- 384. Alibaba's Qwen3.8-Max Writes 7,600 Lines of Code in Five Days
- 385. OpenAI's Astra Solves 10 Long-Standing Math Problems
- 386. Meta AI’s Memory Coach Outperforms Constant Recall for Long Tasks
- 387. AI tools flag thousands of flaws, but few get weaponized
- 388. AI Deletes Spreadsheet Data When Asked to Clean Entry
- 389. AI Coding Agents Speed Tasks but Can't Verify Science
- 390. Chinese AI Researchers Turn to X for Technical Audience
- 391. Google DeepMind's Gemini AI now controls entire humanoid robots
- 392. Apple CEO Tim Cook Suggests Possible Paid iCloud Tier for AI Features
- 393. Google DeepMind Demos AI Orchestrating Boston Dynamics Spot Robot
- 394. Ex-OpenAI Researcher Sees USD 100 Billion Bet on Training Data Beyond Scaling
- 395. Hugging Face breach shows "reasonable measures" amid noisy OpenAI hack
- 396. OpenAI Restricted AI Model Access After Hugging Face Breach
- 397. 2025 Study Finds AI Builds Trust Faster Than Human Scammers
- 398. Nimble's New Web Search Agents Cut AI Token Costs by Half
- 399. DeepMind AlphaFold team disbands as researchers depart for Anthropic, Isomorphic
- 400. Hugging Face Traces 17,600 Actions by Compromised AI Models